Libertarianism in disguise

There is much to agree with, and to admire, in Raymond Tallis’s new book. It presents a fascinating and unusual defence of the reality of human freedom, a defence which Tallis himself describes as a version of compatibilism (though I shall argue below that that is a rather misleading way of putting the claim I think he actually wishes to make). It offers an extremely original set of reflections on the concept of causation, casting scepticism on its conceptualisation as a kind of constraining force which might negate, or count as some sort of obstacle to, our own agency. Unlike most work on the free will problem, it is a book which takes account not only of much recent work in the analytical tradition, but also a number of important ideas from the work of continental philosophers such as Heidegger and Sartre. It is beautifully written and full of interesting examples. In what follows, though, I shall focus on my points of disagreement with Tallis, out of a conviction that it is in this way that progress on the issues where we differ is most likely to be made. Like Tallis, I believe the case for human free will is incontrovertible. But unlike Tallis, I am sceptical that it is possible to turn the trick of offering a satisfactory defence of the existence of free will without questioning the idea that to all intents and purposes, the world into which we human beings and our actions are supposed to be fitted is a deterministic one. I am also sceptical, as I shall shortly make clear, of Tallis’s apparent conviction that he does not himself hold views which are, in the end, inconsistent with the truth of determinism. The picture he eventually paints, of a humanity whose intentional, thought-suffused and co-operative modes of activity and interactivity transcend and lie apart from nature, seems to me in fact to be a variety of libertarianism – that is, it is a view which ultimately defends a free will inconsistent with the view that our world (which of course includes we humans) is wholly deterministic.

What exactly is compatibilism about free will? Usually, the compatibility in question is taken to be compatibility with determinism. In his prefatory note, Tallis defines determinism in a fairly standard way as the view that “only one course of events is possible: the state of the world at any given time fixes its state at future times” (Tallis, 2021, xiii), but he then goes on to assert that his own view counts as a form of compatibilism because it does not deny “a key aspect of determinism; namely, that the natural world unfolds according to unbroken – and hence seemingly unbreakable – habits that science unpacks as laws” (Tallis, 2021, xiii-xiv). But even though this idea is a key aspect of determinism, ^[1] it is certainly not the whole of it. Determinism asserts not only that the natural world unfolds according to unbreakable laws, but also that there is nothing, no fact, thing or aspect of our present or future reality which has not been entirely settled only by those unfolding laws (in conjunction with the actual circumstances which reigned at previous stages of the universe). The mere thesis that there are unbreakable laws – even that there are a very large number of such laws – is hence a far less exigent thesis than determinism itself. But this in turn means that it is far easier for a position to be compatible with the former thesis than it is for a position to be compatible with the latter. The overall result of this situation, in my view, is that the thesis Tallis calls ‘compatibilism’ has not really earned the right to be regarded as compatibilism as that is usually understood, because it is not (apparently) explicitly committed, once one looks closely, to the compatibility claim which is usually the definitive commitment of compatibilists. Rather, Tallis’s compatibilism is a view to which I imagine even many traditional libertarians would be perfectly happy to sign up. For contrary to what is sometimes claimed by their confused opponents, there is no need at all for libertarians to deny the existence of unbreakable laws – all they need deny is that those laws alone are entirely sufficient, in and of themselves, to dictate a single future from any given actual point in time.

It is unclear whether Tallis himself is confused about the relationship between the existence of unbreakable laws and determinism – the careful reference mentioned above to “a key aspect of determinism” suggests he is not. But since confusion on this score is very prevalent in the literature, it may be worth saying a little more to defend the idea that the laws are unbreakable by no means entails the idea that the laws determine everything that happen. The point can be seen instantaneously by considering a very simple ‘toy’ example. Consider a world in which there is just a single law – say, that anything that is red at any instance in that world must turn green 10 seconds later and then back to red, 10 seconds after that. ^[2] Everything else in this world, let us suppose, is sheer chaos – no further patterns of any significance may be observed. This, then, is a world in which all the laws there are are ‘unbreakable’ laws. But clearly the world I have described is a non-deterministic world – one which is indeed absolutely constrained by its single law, but where much nevertheless happens that is not dictated by it.

I would not wish to disagree with Tallis that the natural world does indeed unfold according to laws which are unbreakable (even if we might not, in certain cases, quite have arrived yet at the best possible formulation of those laws). Objects are attracted to one another in accordance with the laws of gravity. The temperatures, pressures and volumes of gases (to take an example much used by Tallis himself) are always related in ways we can codify quite simply. Quantum systems behave according to state vectors which govern their evolution over time. And so on. But to accept that all this (and much more) is true is still to accept very much less than the thesis of determinism. ^[3] The idea that the natural world unfolds according to unbreakable laws simply does not entail that only one course of events is possible – it entails only that whatever courses of events are possible must ‘obey’ those unbreakable laws. But the question is why there might not be many such futures that do so. What is the reason for supposing the laws whittle those futures right down to one?

Compatibilism about the existence of unbreakable laws with the claim that we have free will is therefore a very much more reasonable position than is compatibilism about determinism and free will. The former might plausibly be regarded as a minimum requirement for the scientific sanity of any attempt to defend free will – because we have such strong scientific grounds for supposing that such laws exist. But what are our grounds for believing in determinism? I firmly believe that none have ever been put forward that can bear more than the slightest scrutiny. If I am right about that, there is simply no need for a defender of free will to be compatibilist about free will and determinism – and a good job too, for most such attempts usually only succeed by way of watering down the conception of free will on offer, in order to earn the right to the assertion that free will is available within the constraints of a determined world. But there is no sign that Tallis wishes to do any such watering down – and no sign, therefore, that he actually has any desire to be a compatibilist as that doctrine is usually conceived, nor indeed any resources available to make the position look tenable. That Tallis’s so-called compatibilism is more reasonable than the garden variety is of course a good thing for the view – but it is a bad thing for its claim to be reckoned a variety of compatibilism in the first place.

Tallis’s overall account of how agents fit into the world, indeed, seems dependent upon a number of moves which seek to undermine the models and metaphors which undergird deterministic conceptions of the universe – so it is particularly curious that he is insistent on calling his position ‘compatibilism’. His account of laws, for example, seeks to emphasise their status as human creations, whose central concepts and variables reflect human interests and concerns. He makes an interesting distinction, on which he places much stress, between what he calls the ‘habits of nature’, things “which we must assume have not changed, or not at least in the short time since human beings first became scientists” (Tallis, 2021, p. 29) and the ‘laws of science’ which do change and which must therefore “belong to a virtual space outside of nature, occupied by humanity” (Tallis, 2021, p. 29). Laws, on his conception of things, are scientific products, bearing the stamp of the interests and concerns of the creatures who formulated them – whereas the ‘habits of nature’ are objective realities, ways in which the world always has behaved – and in which, we may assume, it will continue to behave. But Tallis’s reason for making this distinction seems largely to be to exploit it in order to argue that our capacity to encode our understandings of the habits of nature by means of transformation into scientific laws “proves that we are not entirely subject to the former” (Tallis, 2021, p. 50) (i.e. to the habits of nature). And this conclusion confused me. It seems to suggest that we humans escape what would otherwise be complete ‘subjection’ to the habits of nature – and this is not a claim one normally associates with defenders of compatibilism. That this is indeed Tallis’s view is confirmed, seemingly, by the conclusion of Chapter 2, where Tallis writes this:

[W]hile knowing a law does not entail any exemption from the habits of nature, the complex, deliberate processes by which we come to know the law indicates a relationship to nature that is remote from the subservience that would flow from total engulfment in the natural world (Tallis, 2021, p. 58).

These do not sound very much like the convictions of a compatibilist to me. If we are not ‘totally engulfed’ in the natural world, while nevertheless not being exempt from the ‘habits of nature’, what can that mean other than that our actions are not subject to determination (but at best to constraint) by those habits? Tallis’s vision is of agency as something which introduces into nature ways of shaping and producing the future which amount to something far more than the mere entrenchment of a range of perhaps more complex ‘habits’ for nature to conform to. How, then, is the conclusion to be avoided that Tallis is in fact a libertarian? – someone who believes that the world is (overall) indeterministic, precisely because (at least in part) of the existence of agency within it?

I therefore believe that Tallis may be mistaken about the nature of the position for which he has argued (or at the very least, that he has labelled it with a misleading term). But what about the position itself? Though I would myself endorse and applaud many aspects of the libertarian conclusion to which Tallis is led, I confess to being somewhat unclear quite how the arguments for it are supposed to go – and also uneasy about what can sometimes feel to be its anti-naturalistic stance concerning the place of humanity in the grand scheme of things. I shall comment briefly on each of these separate concerns in turn.

Arguments first: what are Tallis’s arguments for thinking that we have a variety of free will which permits us to escape ‘engulfment’ by the habits of nature? There is more than one such argument, but for reasons of space I shall focus here on the line of thinking that is developed in the chapter on laws. Let us re-state the central thought, which we have already met – it is that our impressive capacity to investigate our world in such a way as to produce the laws of science somehow proves our freedom from that total ‘engulfment’. But how does this proof go, exactly? Why should it be ruled out that we are perfectly well able to discover and perhaps even develop some kind of understanding of some of the regularities (“habits of nature”) to which we nevertheless are entirely subject? – and to encapsulate the principles governing some of them in the form of scientific laws? Might we not be capable of the generation of scientific knowledge in all the abstract splendour we know it to display and yet still have our behaviour entirely dictated by the underlying habits of nature which those laws might seek to codify? If not, why not? The point is presumably not just that the astoundingly creative and spontaneous-seeming processes underlying collaborative scientific progress, when we really get them properly into view, are ultimately impossible to conceive of as being generated by a population of creatures each of whose thoughts and motions is no more than the complex upshot of a set of deterministic interactions. Actually, I have much sympathy with that general thought – but all the same, it is presumably not the kind of point that is likely to move any card-carrying compatibilist determinist. No compatibilist worth their salt has ever doubted the consistency of determinism with all the known facts about social collaboration, human ingenuity and creativity, conceptual thought, and so on. We must therefore seek a more complex conception of how the argument is supposed to go.

Tallis makes a number of very interesting points in his chapter on laws which seem intended to help us free ourselves of the idea that laws might be the sorts of things that might keep us under a kind of subjection that would be inconsistent with our freedom. One central idea is that we are deluded if we suppose that there is such a thing – even in principle – as the ‘correct’ list of all laws, perfectly formulated in terms of some utterly neutral, scale-free and interest-free conceptual scheme. Tallis is at pains to emphasise the extent to which the laws as we have them are the product of the contingent course that the history of science has followed – shaped by the interests of the scientist, inherited techniques of investigation, available equipment, and so on. Tallis is moreover sceptical about the idea that it is simply a matter of time before we escape the limitations of these contingencies:

The key point is that what is picked out by the law of science . . . is not inherent in nature as its intrinsic habits or master-habit. This is not just a temporary state of affairs, that will obtain until we arrive at a 100 per cent accurate complete account of underlying principles or a single principle governing change when the laws of science will be identical with the habits of nature. (Tallis, 2021, p. 47)

But even if we accept Tallis’s view that the laws of science are indeed (and will always be) products of such things as human history, culture, biology and happenstance, the question is what we are to say about the habits of nature themselves. Might nature, acting in accordance with those habits (for example, in the ‘natural’ electrical and chemical processes in our brains) not in fact produce the very things we think of as being our free actions? How can we know it is not so? Tallis appears to rule the possibility out, asking at one point how, if human consciousness were entirely constructed out of neural activity, it could “stand outside of nature sufficiently to discover that this capacity to stand outside of nature is located in neural activity entirely subordinate to the laws of nature” (Tallis, 2021, p. 48). But that is a rhetorical question – and the neurodeterminist is likely to reject the implication that we can see any ‘in principle’ difficulty in the case. And I do not in fact myself see what the in principle difficulty is supposed to be. Whatever kind of ‘standing outside nature’ we might need to do in order to be scientists, can we not be confident, in view of our evolved status, that it is not, at any rate, any kind of standing outside nature that is inconsistent with being very firmly part of it? And if we are firmly part of nature, what exactly is it that prevents our complex actions from being merely the products of its habits, even as we move through the thought processes by means of which we discover them? I do not believe that our actions are nothing more than the manifestations of underlying deterministic processes any more than Tallis does – but I fail to see how our capacity to investigate nature in such a way as to produce the laws of science offers any kind of proof of non-engulfment.

Now for the second concern: anti-naturalism. Tallis’s book treads a fine line between the secular humanist position which he avows, always respectful of science in general, and of evolutionary theory in particular, and the human exceptionalism which he eventually defends. It might be said that there is nothing intrinsic to human exceptionalism that is necessarily in conflict either with science or with evolutionary theory, and I would agree (see below). But the details of the exceptionalism matter. And there are places where Tallis says things with a somewhat mystical cast which make me uneasy – such as that actions “have their roots in something outside of the natural world” (Tallis, 2021, p. 102). What he seems to mean by this is that actions depend upon the capacity to imagine various possible outcomes and hence on a capacity to engage not merely with the actual world, but also with a range of possibilia. But it seems to me to be an exaggeration of this undisputed fact to say that this requires actions to have their roots in something outside of the natural world. I would be inclined, rather, to say that their roots lie not so much in the possibilia themselves, as in our capacity to imagine them. And surely these abilities are made possible by thoroughly natural developments, doubtless aided and indeed bootstrapped to new heights by cultural innovations such as language, calendars, art, etc. – but nevertheless rooted in naturally generated capacities, more primitive versions of which we might therefore perhaps expect to find in some of our animal cousins.

This brings me, finally, to an aspect of Tallis’s book on which I should certainly comment, since he devotes an Appendix to discussing my own views on the matter: that is, the relation between human and animal agency. Tallis believes that there is “a huge gap between ourselves and other living creatures, even our nearest primate kin” (Tallis, 2021, p. 198). I do not in fact want to dispute this claim – of course there is such a huge gap. How could it reasonably be doubted that there is such a gap, in light of our advanced linguistic capacities, scientific knowledge, artistic prowess, technological abilities, etc.? What I doubt – and what Tallis perhaps does not doubt – is only that the gap is such as to consign the behaviour of non-human animals entirely to the realm of the non-agential. It is not clear to me whether Tallis would really disagree – he does note that freedom is a matter of degree, which perhaps is intended to permit the admission of a certain measure of it to some nonhuman animals. On the other hand, he does also say that “Beasts seem to pinball through their lives” (Tallis, 2021, p. 198) and that “[T]he behaviour that serves the fundamental Darwinian purpose of survival and reproduction can be plausibly boiled down to conditioned and unconditioned reflexes, instincts and tropisms, tuned by passive associative learning to the contingencies of a particular environment” (Tallis, 2021, p. 198) – and I do doubt these claims. Animal agency cannot all be dismissed as mere ‘pinballing’. ^[4] Watching New Caledonian crows resolve the complex puzzle of how to use a range of tools (some required in order to release others necessary to the task) in order eventually to obtain some food looks very hard to explain as produced by the mere operation of such things as reflexes, tropisms and instincts – and more also than the product of strictly associative learning, since it requires putting things learned together in new and innovative ways. It seems to demand, indeed, the very kind of envisaging of future possibilities (albeit short-term and local ones) which are crucial to Tallis’s conception of the essence of agency.

In short, then, I am not at all doubtful that cultural phenomena of numerous kinds make some version of human exceptionalism quite unavoidable – but I am not convinced that Tallis has quite landed on the right version. If we are properly to understand free will, it will be essential to maintain in view both the sophisticated manifestations of the phenomenon which are perhaps most impressively displayed, as Tallis stresses, in the joint enterprises of humanity, and also the more lowly physical freedoms inherent in animal nature and on which such sophistication must ultimately depend.